Analyzing and segmenting customers based on their value can often be challenging. In fact, the difficulty lies in identifying what is meant by value. In fact, it is not obvious that the value is defined only by how much a user “has billed”. In fact, it may have “invoiced” a lot but its purchase was made so long ago that it can no longer be considered.

To assign a value to our customers that takes into account the main factors, it is possible to use a method that is now tried and tested in the marketing field, the RFM Matrix.

The three variables of the RFM model for a user are:

• Recency: How long has it been since your last purchase?
• Frequency: How often did you buy?
• Monetary: How much did you spend?

Blendee, by default, bases the calculation of the variables on a period of 365 days, this value can be modified by agreement with the Blendee Team.

Having said about the period taken into consideration for the calculation, let’s now see how Blendee specifically calculates each value of the Matrix.

Recency

To calculate “how recently” customers have purchased, you must first define an index that identifies whether a user purchased a long time ago or more recently.

To create such an index, Blendee will analyze the dates of the last purchases made by each customer and based on this it will proceed to create the reference index where, a user who has purchased a long time ago, will be assigned the value 1 while a user who has recently purchased will be assigned the value 5.

Frequency

Also to calculate “how frequently” a customer has purchased Blendee first proceeds to create an index based on all purchases made by individual customers in a year. We will then have users who have placed 10 orders, 20 orders or 4 orders, and so on.

Blendee, given all purchase frequencies, will define the number of users who have made a large number of purchases, for example 10 or more orders, and will assign a value between 1 and 5 where 1 is a low frequency of purchases and 5 a high frequency of purchases.

Monetary

Surely Monetary is the easiest value to understand. Once again, Blendee will take into account the value of purchases made by each user in a year and based on this it will create indices that will allow you to assign a value from 1 to 5 where 1 is a low purchase value and 5 a high purchase value out of all the purchase values made by each customer.

Having defined the working mechanism of the RFM Matrix in Blendee, let’s try to see a case:

The user John Smith has an RFM index of 145: starting from the definitions, we can say that the user has made purchases in the distant past that is greater than the total average of other customers. However, we can see that these purchases were repeated to a greater extent by the majority of users and that these purchases also brought a high value compared to the expenditure made by the majority of customers.

The analysis therefore suggests that the user, after having been a very important customer for us in the past, has stopped making purchases and that therefore all the necessary activities should be put in place to recover him.

The preferences regarding the time over which the RFM index will be calculated can be changed from the Settings -> Advanced Settings menu. By default, the values set are: 365 days on which to perform the RFM calculation, 24 hours for the duration of the cart and 4 hours for the abandoned cart.